7458 (4) Chemical evidence supporting the formation of a bicyclo[2.1 .O]pentanone derivative in the vinylogous Wolff rearrangement is available: Doering and Pomerantz prepared the highly strained cyclobutanone (ii),albeit in low yield, via the copper catalyzed decomposition of diazoketone (i): W. von E. Doering and M. Pomerantz, Tetrahedron Lett., 961 (1964):
'8'
COCHNZ i
ii
More recently Zimmerman and Little reported that when the copper catalyzed decomposition of iii was carried out in the absence of methanol an intermediate, as yet unidentified, is formed. This intermediate displays an infrared band at 5.60 p (1785 cm-'). Upon methanolysis ester iv is formed in addition to a variety of other products: H. E. Zimmerman and R. D. Little, J. Am. Chem. SOC.,9 6 , 4 6 2 3 (1974).
'"E
COOCHj
Ph 111
IV
(5) Thisnew compound was purified by preparative vapor phase chromatography (VPC) and then gave satisfactory elemental analysis and/or exact bass of the molecular ion by high resolution mass spectrometry. (6) M. W. Rathke, J. Am. Chem. SOC., 92, 3222 (1970). (7) M. W. Rathke and D. Sullivan, Tetrahedron Lett., 4249 (1972). (8) J. L. Herrmann, G. R. Kieczykowski, and R. H. Schiessinger, Tetrahedron Lett., 2433 (1973). (9) All yields were determined by VPC calibration and were not maximized. (10) B. M. Trost and P. E. Strege, J. Am. Chem. SOC.,97, 2534 (1975). We are grateful to Professor Barry M. Trost for supplying these data. ( I 1) R. Pappo, D., S.Allen, Jr., R. U. Lemieux and W. S. Johnson, J. Org. Chem., 21, 478 (1956); for oxidative cleavage of conjugated ketones see D. M. Piatak, H. B. Bhat, and E. Caspi, J. Org. Chem., 34, 112 (1969). (12) For a recent review see B. C. Mayo, 0.Rev., Chem. SOC.,49 (1973). (13) J. A. Marshall and D. J. Schaeffer, J. Org. Chem., 30, 3642 (1965): and D. Caine and F. N. Tuller, J. Org. Chem., 34, 222 (1969). We are grateful to Professor Caine for the generous sample of enone 10. (14) A. Segre and J. I. Musher. J. Am. Chem. SOC.,89, 706 (1967); D. K. Dalling and D. M. Grant, ibid., 89, 6612 (1967). (15) S. Wolff and W. C. Agosta, J. Org. Chem., 38, 1694 (1973). (16) L. Horner and E. Spietschka, Chem. Ber., 88, 934 (1955). (17) Comparison here was by IR, NMR, and VPC retention data. (18) Similar stereoseiectivity, leading to tricyclic ketones Ila-b and Iila-b (e.g., 9: 1, respectively), was observed in the copper catalyzed decomposition of y,&unsaturated diazoketones (la-b); M. Mongrain, J. Lafontaine, A. Belanger, and P. Deslongchamps, Can. J. Chem.. 48, 3273 (1970): also see P. M. McCurry. Jr., Tetrahedron Lett., 1845 (1971).
I
1
'3 TIME
I
20
3
(minutes)
Figure 1. Dependence of the Cr(NH1)jNH2Cl3+ quantum yield on irradiation time; initial concentration of [Cr(NH3)jN3](C104)2 is 2.8 X M in 0.3 M HC1, l o einstein I.-' min-I.
bilization of the nitrene intermediates by additional d a - p a bonding.Ic As yet, however, nitrene intermediates have not been detected in the photochemical reactions of first transition series azidopentaammine complexes. Analogous to rhodium(II1) and iridium(II1) complexes, the +2 oxidation state of acidopentaamminechromium(II1) is less favorable. Moreover, previous investigations of thermal substitution reactions of Cr(NH3)5N32+ have indicated that the Cr-N bond is unusually table.^ With these observations in mind, we have reexamined the 31 3-nm photolysis of Cr(NH3)5N32+ to determine if nitrene intermediates occur. Previous investigations of the ligand-field photosensitivity of C ~ ( N H ~ ) S have N ~ ~established + aquation of ammonia as the predominant p r o c e ~ swhereas ,~ irradiation in the L M C T region, X 5 3 3 0 nm, resulted in a general decline in the ultraviolet absorption spectrum and the evolution of N2 gas.5 When acidic or neutral solutions of C ~ ( N H ~ ) S are N ~exposed ~+ to 3 13-nm radiatioq6 the quantum yield of dissappearance of I I1 UI a, R = H C ~ ( N H ~ ) S Ndetermined ~~+, from the decrease in absorbance b, R=OCH,CH.O a t 280 nm and/or 265 nm, was 0.48; in excellent agreement with the previous determination of 0.45.5Also, azide ion was Amos B. Smith, III,* Bruce H. Toder, Stephen J. Branca not detected in the p h ~ t o l y t eContrary .~ to the previous report,5 The Department of Chemistry and however, the inability of the photolyte to reduce C O ( N H ~ ) ~ ~ + , The Monell Chemical Senses Center Co(NH3)5ClZ+,or Co(NH3)5Br2+ and flash photolysis studies University of Pennsylvania indicate photoredox decomposition is not a major reaction Philadelphia, Pennsylvania 19174 pathway. When degassed M Cr(NH3)5N32+ solutions Received May 28, 1976 containing to M N a I were exposed to a filtered (lo-* M NaI) 210-5 flash, 20% of the complex was decomposed and N2 was formed, yet no transient absorbance of 12The Photochemical Reactions of was o b ~ e r v e d . ~ Although ~ , ~ ~ , * the absence of an 12- absorbance Azidopentaamminechromium(II1). Evidence for a First indicates photoredox does not occur, its absence was further Transition Series Coordinated Nitrene Intermediate confirmed by flash photolysis. A degassed M Cr(NH3)5N32+ solution was exposed to a filtered (lo-* M Sir: NaI) 250-5 flash and analyzed a t 740 nm, the absorption The recent discovery by Basolo and co-workers of an effimaximum of C r ( H ~ 0 ) 6 ~Sprectra +.~ recorded before and after cient photoinduced reaction mode involving nitrenes' has rethe flash indicated 50% of the complex was decomposed by the vived considerable interest in the photochemical reactions of flash, but no transient absorbance was observed. Assuming a azido complexes of transition m e t a k 2 In contrast to the redox minimum signal to noise ratio of 2:1, the absence of a n abmode observed in the photolysis of C O ( N H ~ ) ~ Nthe ~ ~rho+ , ~ sorbance establishes that photoredox accounts for less than dium(II1) and iridium(II1) analogues reacted principally via 12% of the overall photochemical reaction. a coordinated nitrene, M(NH3)5NH3+, pathway.' The difThe previous study which proposed a photoredox mechanism ference in reaction modes with the heavier metals was attribreported the yield of N2 to be essentially the expected 1.5 mol uted to less favorable +2 oxidation states and increased staof N2 per mole of complex decomposed. However, we find the Journal of the American Chemical Society
/ 98:23 / Nouember 10, 1976
7459
yield of N2 to be less than 1.5 mol of N2 per mole of complex decomposed. The amount of gas liberated during the photolysis was quantitated by gas chromatography. A M [Cr(NH3)5N3](C104)2 solution containing 0.3 M HC1 was degassed by repeated freeze-thaw cycles. The spectrum of the solution was recorded before and after photolysis and the number of moles of Cr(NH3)5N3*+ decomposed was determined. The noncondensable gas (at liquid NZ temperature) was quantitatively transferred by a Toepler pump to a Gow Mac gas chromatograph equipped with a 6 ft X 0.25 in. Porapak Q column at room temperature. The GC trace indicated N2 was the only gaseous product and a series of five experiments yielded 1.1 f 0.2 mol of N2 per mole of C ~ ( N H ~ ) S N ~ ~ + decomposed. The apparent excess of nitrogen gas may be due IC 20 to an inability to account for a possible absorption of the product(s) in determining the number of moles of complex Y[HC decomposed and/or secondary photolysis1aor thermal instaFigure 2. Dependence of the reciprocal of the Cr(NH3)5NH2CI3+ quanbility of the primary product of the photochemical reaction (see tum yield on the reciprocal of the HCI concentration. Initial concentration below). M. of [ C T ( N H ~ ) ~ N ~ ] ( C I is O& 3X When M [Cr(NH3)5N3](C104)2 solutions containing 1 .O M HC104 are exposed to 3 13-nm radiation nitrogen is agent is the observation that only Co(NH3)5H203+ is reduced evolved, but the photolyte does not oxidize iodide ion. In 1.O whereas C O ( N H ~ ) ~Co(NH3)5C12+, ~+, and Co(NH3)5Br2+ M HC1, however, 313-nm photolysis of a M Cr(NH3) are not reduced. A number of pieces of evidence, however, 5N32+ yielded a photolyte which oxidized iodide to iodine with suggest that the reductant may be a nitrene product such as a quantum yield of 0.23. These results indicate that the initial Cr(NH3)5NH20H3+ or its conjugate base, Cr(NH3)5photochemical product is a chromium nitrene intermediate, NHOH2+. For example, the yield of Co(I1) decreases signifCr(NH3)5NH3+, which in the presence of HC1 is trapped to icantly as the concentration of HC1 increases. In the presence form the chloramine, known to oxidize iodide to i0dine.l As of Co(NH3)5HzO3+,flash photolysis of I T ( N H ~ ) ~ which N~~+ indicated by Figure 1, secondary photolysis of Cr(NH3)sreacts exclusively via a nitrene intermediate results in a much NH*Cl3+occurs extensively and, as mentioned above, accounts longer lived transient, t 1 / 2 0.1 s, than in the absence of in part for the high yield of nitrogen relative to the amount of Co(NH3)5H203+, t ~ =p 50 ms.Ib complex decomposed.la Additional experiments also indicate A priori, a nitrene reaction pathway, cleavage of an N-N2 Cr(NH3)5NH2C13+ is thermally unstable. A M bond, is thought to originate in an azide centered excited Cr(NH3)5N3*+ solution containing 1 .O M HC1 was photolyzed state.lb3l0 Unlike Co(NH3)5N3*+, however, which undergoes to 25% consumption of the starting material. The photolyte was photoreduction like other acidopentaamminecobalt(II1) then stored in the dark a t 25 'C and aliquots taken periodically complexes, the nitrene reaction pathway of Cr(NH3)5N3*+ were tested with iodide ion. After 20 min, the ability of the is quite different from other acidopentaamminechromium(II1) photolyte to oxidize iodide to iodine had decreased by 20%. complexes.12 In the cobalt(II1) complex, the L M C T state is This thermal instability precludes the use of conventional lower in energy than the azide centered excited state. The ion-exchange chromatography to isolate the Cr(NH3)5difference in behavior of the complexes might then be attribNH2C13+. uted to an efficient coupling of the L M C T state and the azide Flash photolysis experiments as well as the inability of the excited state of the cobalt complex leading to a lower energy photolyte to reduce other acidopentaamminecobalt(II1) photoredox pathway.8 With C ~ ( N H ~ ) S Nhowever, ~ ~ + , the complexes relegate to a minor role the previously proposed azide centered excited state may lie lower in energy than the photoredox process. Moreover, the dependence of the yield of L M C T and M L C T states. An efficient coupling is less likely Cr(NH3)5NH2C13+ on the concentration of HCl, Figure 2, and the isolated azide excited state would then give rise to extrapolates to a limiting yield of 0.5. The close agreement exclusively a nitrene reaction pathway.lo between the limiting yield and the quantum yield of Cr(NH3)5N3*+ decomposition establishes the primary phoAcknowledgment. Financial support of this research from tochemical reaction to be the Research Corporation and The Research Foundation of The City University of New York is gratefully acknowledged. Cr(NH3)sN:' h", Cr(NH3)5N2+ N 2 L ,
-
313 n m
+
Although the M L C T state of Cr(NH3)5N3*+ is expected to be much higher in energy and not important in these experiments, population of this state or one which reacts in a similar manner can give rise to a n interesting non-nitrene reaction forming Cr(NH3)5N2 and N.I0 The fate of the latter species, N , is the formation of NH20H. However, analysis of the photolyte for NH2OH with a-naphthylamine' I indicates this reaction pathway does not occur in these experiments. The initial proposal of a photoredox process for the 3 13-nm photolysis of Cr(NH3)5N3'+ was based, to a significant extent, on the appearance of Co(I1) when a weakly acidic, M HC104, solution of C T ( N H ~ ) ~ N was ~ ~photolyzed + in the presence of Co(NH3)5H2O3+. Although we have confirmed these observations, the mechanism of the reduction is not clear. Inconsistent with the postulation of Cr(I1) as the reducing
References and Notes (1) (a) J. L. Reed, F. Wang, and F. Basolo, J. Am. Chem. Soc.,94,7172 (1972); (b) H. D. Gafney, J. L. Reed, and F. Basolo, ibid., 95,7998 (1973); (c) J. L. Reed, H. D. Gafney, and F. Basolo. ibid., 96, 1363 (1974). (2)(a) G.Ferraudi and J. F. Endicott, lnorg. Chem., 12, 2389 (1973);(b) J. Am. Chem. Soc., 95, 2371 (1973);(c) G. Ferraudi, J. F. Endicott. and J. R. Barber, J. fhys. Chem., 79,630 (1975). (3)A. W. Adamson, Discuss. Faraday SOC., 29, 163 (1960);(b) S.A. Penkett and A. W. Adamson, J. Am. Ct". SOC.,87,2514(1965):(c) J. F.Endicott and M. Z . Hoffman, lbid., 90,4740(1968);(d) J. F. Endicott, M. Z.Hoffman, and L. S.Beres, J. Phys. Chem., 74, 1021 (1970). (4)M. Linhard and W. Berthold, 2. Anorg. Allg. Chem.. 278, 173 (1955). (5) A. Volger, J. Am. Chem. SOC.,93,5912 (1971). (6)The exciting light, a 300-W Xe-Hg lamp, was monochromated to 313 nm, half width 10 nm, with a Bausch and Lomb Model 33-86-26grating monochromator. (7)E. K. Dukes and R. M. Wallace, Anal. Chem., 33,242 (1981). (8)J. F. Endicott, "Concepts of Inorganic Photochemistry', A. W. Adamson and P. Fleischauer, Ed., Wiley, New York. N.Y., 1975,Chapter 3. (9)A. B. P. Lever, "Inorganic Electronic Spectroscopy", Elsevier. New York, N.Y.. 1968,p 289.
Communications to the Editor
7460 (10)J I Zink, Inorg. Chem., 14, 446 (1975). (1 1) (a) F Feigl and V. Anger, "Spot Tests in Inorganic Analysis", 6th ed, Elsewer, New York, N Y , 1972, p 346: (b) M Fradeiro, L Solorzano, and J D H Strickland, Anal. A b s b , 15, 7000 (1968) (12) V Balzani and V Carassiti, "Photochemistry of Coordination Compounds", Academic Press, New York, N.Y , 1970, Chapter 7
I
l A
I
Micheal Katz, Harry D. Gafney* City University of New York Department of Chemistry, Queens College Flushing,New York I 1 367 Received July I , 1976
~
/I
1
11 I
~
- -"---I
rs
HOD
-c____\_&
uo6
4098
Glyoxalase I Enzyme Studies.' 2.* Nuclear Magnetic Resonance Evidence for an Enediol-Proton Transfer Mechanism Sir.
In contrast to the generally accepted 1,2-hydride shift mechanism3 for the action of glyoxalase I, we wish to present evidence that in the conversion of the a-ketohemithiol acetal 2 (methylglyoxalglutathionylhemithiol acetal) to the a-hydroxythiol ester 3 (S-lactoylglutathione) the mechanism involves an enediol-proton transfer rather than a hydride shift. The glyoxalase enzyme system composed of glyoxalase I [ S lactoylglutathione methylglyoxal lyase (isomerizing); E.C. 4.4.1.51, the coenzyme glutathione (GSH), and glyoxalase I1 (S-2-hydroxyacylglutathione hydrolase; E.C. 3.1.2.6) converts methylglyoxal (1) to lactic acid (4). The role of this enzyme system, which is found widely distributed in cells of all forms of life,4.5 in metabolism has become a topic of controversy because of its suggested5 involvement in the regulation of cell division.
0
QH
I
11
CH,C-CHSG 2
0
glyoxalase I
0
II II
CH,C-CH 1
+ GSH
217
9
HQ
I
I1
CH3CH-CSG 3
"7 P
CH,CH--COH
+ GSH
4
Mechanistically, the most intriguing step in the reaction sequence is the rearrangement of the hemithiol acetal 2 to the thiolester 3 for which one might consider either a n intramolecular 1,2-hydride shift or a n enediol-proton transfer mechanism (originally proposed by R a ~ k e r ) . ~Distinguishing .~~,~ the two mechanisms should be Straightforward either by detecting the incorporation (or lack of incorporation) of solvent protons into the product or the retention (or loss to the medium) of an isotope of hydrogen when using a properly labeled substrate. Indeed, studies of the former have been reported in deuterium oxide and tritium-enriched water. In each case the 1,2-hydride shift mechanism was proposed because of the lack of detection of deuterium (3% maximum i n c ~ r p o r a t i o nor )~~ low incorporation of tritium (less than 4%)3bin the lactic acid product, 4. These results, however, do not rule out the possibility of a fast proton transfer mechanism via an enediol taking place in a highly protected active site. As a matter of fact, the detection of any solvent proton incorporation suggests such a possibility. This can be tested by observing incorporation or increased incorporation of solvent protons as the temperature is raised.' W e report here the results of such a study in which incorporation of solvent protons was detected using N M R spectroscopy. At 25 'C there is incorporation of deuterium (ca. Journal of the American Chemical Society
Figure 1. ' H NMR spectra for the glyoxalase I catalyzed conversion of methylglyoxal to lactic acid. Spectra A and B represent the enzyme reaction at 25 and 35 "C, respectively; A' and B' are the corresponding control spectra. Chemical shifts are quoted in parts per million from DSS and the assignments are discussed in the text. The unassigned peaks belong to glutathione.
15%) when the enzyme reaction is run in deuterium oxide. The incorporation increases to ca. 22% when the temperature is raised to 35 'C. These observations clearly demonstrate that
/ 98:23 / Novenrber 10, 1976
